In the realm of customer engagement, the chatbot has fast become many marketers’ go-to automated service. Over 1.4bn people a year now interact with bots — or “virtual digital assistant software” — and their proliferation of use isn’t expected to let up anytime soon. The market for chatbots is expected to reach $7.7bn in value by 2025.
It is not enough for businesses to simply launch a chatbot, however. Companies that want to make the most of the opportunity must understand that the secret to an effective, successful service lies in the data that chatbots use. Data is the tool that allows chatbots to really understand the people they are chatting with, enabling them to respond to any questions posed in a useful and engaging way.
Just as humans need accurate information to learn skills, so too does effective automated technology. Consider, for example, how Amazon had to scrap its artificial intelligence (AI) recruiting tool because the data it was built on had taught it to immediately disregard female applicants. More importantly than this, for automated services such as chatbots to be truly valuable, the data they need must not only be accurate but also accessible instantly.
Under the AI umbrella: Machine learning, natural language and computer vision — key to enhancing chatbot capabilities
Anshuman Singh, head of digital for Mindtree in Europe helps Information Age explore the importance of AI and natural language in developing chatbots. But, be careful to not forget the human element. Read here
The need for speed
The best chatbots are the ones which can converse with users in a human-like way, disguising the automated software that drives them. In order to do this, they need to be able to answer almost any question a consumer may have on a product, service or offer, instantaneously.
Until now, chatbots have predominantly had to rely on scripted responses—to which there are some clear drawbacks. Scripting restricts a bot to set texts of information, inherently limiting the range of questions it can answer. More importantly, people find these scripted chatbots incredibly unsatisfying to use. ZDNet recently found that the winners of the “Second Conversational Intelligence Challenge”, a chatbot competition set up by Facebook and Microsoft, still left a lot to be desired when it came to usability and consistency.
There are clearly still operational and reputational issues for chatbots to overcome, despite their increasing adoption. One answer lies in instantaneous data delivery, which will allow chatbots to answer questions accurately without the need for scripts. Real-time data delivery models, such as those provided by data-as-a-service (DaaS) companies, can ensure a chatbot has the information it needs to be informative and useful. They can provide virtual assistants with the information they need to answer a vast range of questions in a natural, human-like way that is satisfying for the user.
What is a chatbot? How to make a chatbot? Why chatbots fail? All your questions answered
What is a chatbot? How do you make a chatbot? Why chatbots fail? In this article, we answer all these questions with the help of Howard Pull, strategic development director at MullenLowe Profero. Read here
Making data go further
Even though we are only now seeing chatbots become a mainstream customer engagement tool, they have in fact existed in one form or another for two decades. MSN used a chatbot known as SmarterChild in its fondly remembered MSN Messenger app as early as 2000, and the band Radiohead even created a chatbot called “GooglyMinotaur” to promote their 2001 Amnesiac album and tour.
Some modern chatbots are still being used in this way, predominantly as a novelty, promotional tool. For instance, WholeFoods is using chatbots to provide recipes and cooking inspiration while Marvel is using them to promote comic books by simulating conversations with Spiderman. However, there are deeper and more effective ways to genuinely engage customers using bots. These are emerging thanks to real-time delivery models that are empowering the technology to push forward in the field of customer service.
With DaaS allowing instant access to up to the minute data, chatbots can become an always-on tool for customer engagement. Skyscanner, the flight booking website, is one company using chatbots to provide valuable interactions with customers. People using the bot receive instant recommendations on flights across the world, at any time, as well as advice on destinations and price alerts. When used on Skype, the chatbot can even be added into and interacted with in group chats—as if it was an additional member of a holidaymaker’s group.
Of course, recommendations made – and chatbot conversations themselves – will only truly be valuable if there is an understanding of who a user is, knowing their needs, wants and preferences. This is again where DaaS solutions become vital. Chatbots can use AI coupled with DaaS to pull the most relevant existing data on a customer, as and when it is needed. They can then use this comprehensive information to draw together their profile, demographic, likes and dislikes. This can drive a greater number of positive outcomes, by giving consumers exactly the type of service they want.
How chatbots can raise business intelligence, analytics adoption and success rates
Creating the “data loop”
When used correctly, chatbots are not just an impressive customer engagement tool, they can also become an important means for ongoing data analysis. The more customers chat with a bot, the more data they will input. When DaaS is brought into the fray, a customer’s evolving preferences, searches, wants, needs and communication style can be automatically updated.
This creates a cycle of perpetual improvement, or a “data loop”. The information provided to a chatbot from a customer’s interactions can be instantly applied into their data set, meaning that a chatbot is constantly using up-to-date, accurate and clean information. Real-time data updates mean that offers and communication can remain as relevant, personal and engaging to a customer as possible.
Chatbots have a long history of use and in some cases are becoming indistinguishable from customer service agents. However, to continue being effective, and to achieve their potential as service tools, the AI and machine learning “brains” that drive chatbots need to be fuelled by data. As is the case with all customer engagement, the more you know about the person the better the engagement will be, and chatbots give companies the opportunity to learn about their customers from the very information they input.
Customer engagement through chatbots, facilitated by data delivered in real-time, is the future. However, success ultimately depends on two things: having access to enough data to make sensible engagements and decisions, and having data delivered at exactly the right moment. It is this combination of accuracy and speed which will empower chatbots to make customer engagement more seamless than ever.